[目的]研究西安春季首场透雨的马尔科夫链预测方法.[方法]根据透雨出现时间具有相依随机变量的特点,选取西安市1959～2010年7个气象台站的春季(3～5月)降水量资料,采用气象上春季首场透雨标准确定其出现日期,按照研究问题序列的长短和实际情况将其划分为6个状态空间,并采用马尔科夫链建立春季首场透雨预测模型.[结果]应用马尔科夫链预测的2009、2010年西安市春季首场透雨的出现时间与实况相符,预测效果较好,其方法思路清晰,计算简便,且具有一定的可靠性和实用性.[结论]该方法为实际业务预报春季首场透雨提供科学的参考价值.%[Objective] The aim was to study the Markov Chain prediction method of first heavy rain in spring in Xi'an Province. [Method] According to the dependent random variables of the occurrence of heavy rain, precipitation in spring in seven meteorological stations in Xi' an City from 1959 to 2010 was selected. Its occurrence date was determined by the standard of first heavy rain in spring in meteorology. According to the length of the sequence of the problem and actual situation, six states were divided. And by dint of Markov Chain, first heavy rain prediction model in spring was set up. [ Result] The predicted occurrence time of first heavy rain in spring in Xi' an in 2009 and 2010 was consistent with the actual situation. The prediction effect was fine. The method had clear thought and was convenient for calculation, with certain dependence and practicality. [Conclusion] This method provided reference value for the actual forecast of first heavy rain in spring.
展开▼
